Primary hepatocellular carcinoma (PHC), one of the three most common causes of cancer death in the world, is poorly characterized from a genetic epidemiologic perspective. Epidemiologic studies have firmly established the association of chronic hepatitis B virus infection with PHC risk. However, the majority of chronic viral carriers do not develop PHC. Even among life long carriers, the risk of tumor development is less than 25%. Moreover, this risk does not appear to be distributed uniformly throughout the population. Anecdotal accounts of both geographic and familial clustering are common in the literature. Whether this clustering represents familial aggregation of environmental co-factors (e.g., virus infection, exposures, etc.) or genetic susceptibility, has not been examined in a systematic way. Overall, the underlying molecular mechanism responsible for PHC is not known. The possibility of oncogenesis by the introduction or activation of an oncogene has been extensively studied. To date, no persuasive results have emerged. On the other hand, a number of studies have shown genetic changes in tumors consistent with the loss of a tumor suppressor gene. These findings suggest a common etiologic basis of both HBV and non-HBV associated PHC. The overall goal of this research is to understand the genetic epidemiology of PHC. Statistical methods will be used to evaluate jointly the role of genetic and environment factors in families identified through a PHC case. Molecular studies using the biosamples from the index cases will be used to construct molecular karyotypes. These molecular karyotypes will be used to determine whether genetic changes observed in tumors are consistent with loss of a tumor suppressor gene. The genetic constitution of the index cases' normal and tumor tissue will be correlated with their clinical and epidemiologic information in order to evaluate their joint significance. Finally, the genetic profiles will be used to examine the family data for differences in co-variate and aggregation patterns.